Background. Given the limited availability of HLA-matched donors, haploidentical hematopoietic stem cell transplantation (Haplo-HSCT) is increasingly used. Current strategies to deplete T-cells include ex vivo T-cells depletion, incorporation of antithymocyte globulin during conditioning, and administration of post-transplant cyclophosphamide. The aim of this study was to analyze the outcomes of unmanipulated haplo-HSCT using antithymocyte globulin in children. Methods. We retrospectively reviewed the data of 18 children who received an unmanipulated haplo-HSCT for various hematologic diseases or primary immunodeficiency diseases from January 2011 to March 2017. Transplant outcomes and related prognostic factors were analyzed. Results. Patients’ underlying diseases were hematologic malignancies (n=9), severe aplastic anemia (n=3) and primary immunodeficiency diseases (n=6). Median age at haplo-HSCT was 9.2 years and their median follow-up was 12.4 months from transplantation. Twenty-one transplants were performed for 18 patients with 3 patients (17%) re-transplanted for engraftment failure from the same haploidentical donor which resulted in successful engraftment in all. The incidence of acute graft-versus-host disease (GVHD) was 89% for grade II-IV and 28% for grade III-IV. Seven patients (39%) developed chronic GVHD with 4 extensive diseases. No patient developed CMV disease although 13 patients (72%) had CMV antigenemia. Five (28%) patients died from the primary disease, 1 (6%) from GVHD and 1 (6%) from post-transplant lymphoproliferative disease. The 4-y overall survival rate was 57.6%. We developed a scoring system using 3 parameters such as disease status (nonmalignant disease=0, malignancy=1, advanced malignancy=2), the number of previous transplant (none=0, once=1, more than once=2), and the donor gender (male=0, female=1). When patients were grouped according to the sum of risk scores, the OS rates were 100% in the low risk (score 0-1), 66.7% in the intermediate risk (score 2), and 0% in the high risk (score ≥3) group ( P=0.0008). Conclusions. Unmanipulated haplo-HSCT seems a reasonable treatment option for children without an HLA-matched donor. Our scoring system may serve as a useful reference when an unmanipulated haplo-HSCT is considered.